godfreyhe opened a new pull request #13832:
URL: https://github.com/apache/flink/pull/13832
## What is the purpose of the change
*Shuffle is one of the most heavy operation, and many shuffle operations is
redundant, because the input operators have provide the required shuffle. Blink
batch planner has supported removing redundant shuffle through planner rules,
Blink streaming planner could also leverage similar approach to remove
redundant shuffle. Different batch, stream operators require the shuffle keys
and the state keys must be exactly the same, otherwise the state may be not
correct. In this pr, we only support Join and regular Aggregate, other
operators will be supported in the future.*
## Brief change log
- *Do some code cleanup for BatchExecCalcBase and BatchExecCorrelateBase*
- *Introduce join condition equal transfer rules to simplify join
condition*
- *Remove redundant shuffle for StreamExecJoin And
StreamExecGroupAggregate*
## Verifying this change
This change added tests and can be verified as follows:
- *The existing tests could cover the first two commits, e.g.
batch.sql.RemoveShuffleTest, stream.sql.join.JoinReorderTest*
- *Added new test that validates the result of third commit, e.g.
stream.sql.RemoveShuffleTest, RemoveShuffleITCase*
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): (**yes** / no)
- The public API, i.e., is any changed class annotated with
`@Public(Evolving)`: (yes / **no**)
- The serializers: (yes / **no** / don't know)
- The runtime per-record code paths (performance sensitive): (yes / **no**
/ don't know)
- Anything that affects deployment or recovery: JobManager (and its
components), Checkpointing, Kubernetes/Yarn/Mesos, ZooKeeper: (yes / **no** /
don't know)
- The S3 file system connector: (yes / **no** / don't know)
## Documentation
- Does this pull request introduce a new feature? (**yes** / no)
- If yes, how is the feature documented? (not applicable / docs / JavaDocs
/ **not documented**)
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
